Professional Documents
Culture Documents
INTRODUCTION
During the first 1000 days of life, the intestinal eosinophilic esophagitis, infantile colic, inflamma-
microbiota impacts health in later life through tory bowel disease and obesity.7–11
the interdependent development of microbiota, In mice-based studies, AB exposure in the first
immune system, growth and cognitive func- week of life altered microbiota composition and
tion.1 2 After birth, the intestinal microbiota immune function, but gavage with mature untreated
© Author(s) (or their develops rapidly, driven by exposure to microbes microbiota, restored the perturbation and reduced
employer(s)) 2022. Re-use from maternal, environmental and dietary sources.3 the negative health effects.12 To understand the full
permitted under CC BY-NC. No
commercial re-use. See rights During this early development, the intestinal micro- impact in humans and develop restoration strate-
and permissions. Published biota is unstable and susceptible to perturbations gies, more knowledge is needed.
by BMJ. such as those caused by antibiotic (AB) exposure. Worldwide, up to 20% of all neonates are
These perturbations may have long- term conse- prescribed ABs because of (suspected) early-onset
To cite: Van Daele E,
Kamphorst K, Vlieger AM, quences on the developing microbiota and also neonatal sepsis, although in most cases, sepsis is
et al. Arch Dis Child Fetal on the developing immune system,4 growth5 6 and unconfirmed and ABs could be discontinued after
Neonatal Ed have already been associated with increased preva- 48–72 hours.13–15 More prolonged AB exposure can
2022;107:F603–F610. lence of asthma, allergies, coeliac disease, eczema, gradually reduce overall diversity16 and richness17
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;107:F603–F610. doi:10.1136/archdischild-2021-322861 F603
Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/archdischild-2021-322861 on 9 May 2022. Downloaded from http://fn.bmj.com/ on October 28, 2022 at World Health Organisation
Original research
Figure 1 Overview of the collected questionnaires, faecal sampling points and the age categories based on the age (days) at sampling. N, number
of infants for which samples were available for a given time point. d(ays), m(onths) and y(ears) indicate the age category ranging from birth until
around 2 years of age.
of the neonate’s intestinal ecosystem. Additionally, AB type also the first year of life. Nine faecal samples were collected from
determines the microbial perturbation through specific mech- infants (figure 1). Until discharge from the hospital, faeces were
anisms of action and host interactions. Studies in infants have sampled from diapers by hospital staff and immediately frozen
been inconclusive,17–19 but indicated types with faster recovery at −20°C. Sampling continued at home by the parents, using
towards the microbiota composition of controls.20 Therefore, sampling instructions and freezer storage. After 1 year, parents
more comparative studies are needed between durations and delivered the samples to the clinic after transport on ice. A final
types in AB regimes to optimise AB administration.21 sample was taken around 2 years, stored in the home freezer and
In this study, we investigated the microbiota development collected by the study nurse. At the hospital, all samples were
in a subset of the Intestinal Microbiota Composition after AB stored at −20°C.
treatment in early life (INCA) cohort.22 The primary aim was
to investigate the impact of AB exposure in the first days of life
Table 1 Baseline characteristics of the INCA cohort subset included in this study
AB- AB+
AB2 AB7
N 126 56 20 36
F604 Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;107:F603–F610. doi:10.1136/archdischild-2021-322861
Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/archdischild-2021-322861 on 9 May 2022. Downloaded from http://fn.bmj.com/ on October 28, 2022 at World Health Organisation
Original research
normally distributed variables were tested with the Kruskal- reference group. Differences between AB groups were assessed
Wallis test and indicated by their median and IQR. Bonferroni per age interval using ANOVA in the vegan package.40
correction was performed to correct for multiple testing.
RESULTS
Bioinformatic and statistical analysis of the sequence data Baseline characteristics of INCA cohort subset
All analyses were performed in R 3.6.1.27 Samples were stratified The baseline characteristics differed between AB- and AB+
into 11 age-based and right-closed intervals for statistical anal- for gestational age, birth weight and additional AB exposure
ysis (figure 1). The Jenks Natural Breaks Classification (classInt between one to 6 months (table 1). Birth weight z-score (birth
package) was used to calculate the optimal ranges.29 Because of weight corrected for gestational age), however, was comparable.
increasing increments between sampling, the x-axis (age) was AB2 differed from AB7 with regard to additional AB exposure
log2 transformed for visualisation. between 1 and 6 months (5% vs 34%, respectively, p=0.019).
Alpha diversity (within sample diversity) was calculated at ASV Baseline characteristics were comparable between the AB type
level using Picante30 and Microbiome31 packages and following groups (online supplemental table S1).
metrics: Faith’s phylogenetic diversity,32 ASV richness, Shannon
and Inverse Simpson. All except Shannon diversity needed loga-
rithmic transformation to obtain normal distribution for one-
Antibiotic-induced alterations to intestinal microbiota
way analysis of covariance. Consecutive analyses were corrected development
for the baseline characteristic that differed significantly between AB exposure during the first week of life did not alter microbial
AB+ and AB- and between AB2 and AB7. alpha diversity between birth and 2.5 years (online supplemental
Temporal trends in relative abundance were visualised using figure 1). The temporal patterns of the four major phyla (95% of
local regression with locally estimated scatterplot smoothing the average relative abundance) were compared with univariate
using ggplot2. These relative abundances did not meet normality analyses (figure 2). Relative abundance of Proteobacteria was
requirements and were therefore compared using beta regres- high overall, during the first months of life. AB exposure further
sion with BetaReg33 per age interval. The effects of AB on increased this relative abundance at 1 month (mean 43.6% AB+,
31.5% AB-) and 1 year (mean 17.6% AB+, 6.5% AB-). Actino-
Figure 2 Temporal trajectories of the relative abundance of the four main phyla in the developing intestinal microbiota from birth to 2.5 years of
age. Thick lines represent the average with shading showing the 95% CI. Differences between AB+ and AB- were calculated using beta regression for
each age category. * and vertical grey shading: difference in relative abundance (p<0.05), AB+, infants who received AB during their first week of life,
AB-, infants who did not receive AB during their first week of life, LOESS, locally estimated scatterplot smoothing.
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;107:F603–F610. doi:10.1136/archdischild-2021-322861 F605
Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/archdischild-2021-322861 on 9 May 2022. Downloaded from http://fn.bmj.com/ on October 28, 2022 at World Health Organisation
Original research
Figure 3 WU-based PRC analysis is a special case of RDA for multivariate responses in a repeated observation design, which is applied here on the
differed from AB- baseline (figure 3B and online supplemental associated with decreased Bifidobacterium relative abundance
figure 2B). in FF (1 month) (figure 4C), which occurred later than in BF
AB types had a different impact on the microbiota (online infants (2 weeks) (figure 4B). In turn, AB+ was associated with
supplemental figure 3A,B) with ABAMX not deviating from AB- increased relative abundance of Parabacteroides in BF, and with
baseline and ABPEN deviating at 1 week and 1 year with increased increased relative abundance of Enterobacteriaceae and Entero-
relative abundance of Enterobacteriaceae members in WU and coccus in FF infants.
UU. UU-based deviations between ABAMC and AB- baseline were
limited to week one and involved different ASVs compared with DISCUSSION
ABPEN. ABAMC affected WU in the long- term with increased In this prospective, observational INCA study, we examined the
Enterobacteriaceae and Enterococcaceae at 2 weeks and 1 month microbiota development after AB exposure during the first week
and also Bifidobacterium at 1 week and 2.5 years. of life and found perturbations in the faecal microbiota devel-
opment from 1 week until 1 year of age. These perturbations
Impact of delivery and feeding mode on AB-associated included decreased relative abundance of Bifidobacteriaceae
deviations in the faecal microbiota development while potentially pathogenic Enterobacteriaceae increased. This
Due to the relatively low number of faecal samples in the first study adds new insights into long-term compositional shifts after
week per feeding and delivery type (figure 1), effects were only neonatal AB exposure.16 41 Our results corroborate findings in
reported in samples collected between 1 week and 2.5 years. older infants with increased Enterobacteriaceae and decreased
Within AB-, microbiota deviated based on delivery mode from Bifidobacteriaceae after AB administration.16 20 42–44 Importantly,
1 week until 1 month (figure 4D). In vaginally delivered infants, the severity and duration of AB-mediated microbiota perturba-
the AB effect on the microbiota was still significant at 1 year tions increased with longer AB administration (5–7 vs 2–3 days).
with an increase of several Enterobacteriaceae, Enterococca- The results also align with a small study in preterm infants,
ceae and Streptococcaceae and decrease in Bifidobacterium and where >5 days AB exposure intensified perturbations compared
Escherichia-Shigella. In contrast, no AB- mediated deviations with 2–3 days.45
were noted between C-section born infants. Bifidobacteriaceae form a cornerstone in the early devel-
Compared with AB- BF baseline, AB+BF infants only deviated opment of the immune system. They were shown to promote
at 2 weeks, whereas AB+FF infants showed longer deviations B-cell maturation and associations with decreased inflammatory
from the first week up to 1 month (figure 4A). Because there responses and T-regulatory cell acquisition.46 Enterobacteriaceae,
was also a feeding effect during the first 6 months, the AB effect on the other hand, produce toxins and have lipopolysaccharides
was also analysed within the separate feeding groups. AB+ was on their outer membranes, which causes inflammation.47–49
F606 Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;107:F603–F610. doi:10.1136/archdischild-2021-322861
Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/archdischild-2021-322861 on 9 May 2022. Downloaded from http://fn.bmj.com/ on October 28, 2022 at World Health Organisation
Original research
Therefore, it is not surprising that reduced Bifidobacteriaceae, The long-term microbiota effect of ABs and its associated
often combined with an increase in potentially pathogenic negative health outcomes reinforce the need for implementing
Enterobacteriaceae, like Shigella, Klebsiella and Enterobacter, AB stewardship programs53–55 to avoid AB overuse.21 56 57 The
have been associated with immune- mediated disorders like microbiota perturbations were only significant after 5–7 days
asthma.50 Similar deviations were also found in functional disor- compared with 2–3 days AB which could explain previous find-
ders like infantile colic51 and irritable bowel syndrome.52 ings from the INCA cohort: namely higher incidence of infantile
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;107:F603–F610. doi:10.1136/archdischild-2021-322861 F607
Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/archdischild-2021-322861 on 9 May 2022. Downloaded from http://fn.bmj.com/ on October 28, 2022 at World Health Organisation
Original research
colic, wheezing and food allergies in infants exposed for 5–7 days, represent relevant members of the intestinal microbiota such as
but not for 2–3 days.10 58 If this observed difference between 2–3 bifidobacteria more accurately, which makes them especially fit
days and 5–7 days exposure is the result of longer AB exposure or for analysing the intestinal microbiota of infants.23 For future
the result of a concomitant infection or inflammatory response is research, whole genome shotgun sequencing could be used to
yet unclear. The AB7 infants were treated because of suspected increase the accuracy of species and strain detection.73 74 Another
early onset sepsis (EOS). EOS is rare in term infants,59–61 but it limitation may be that environmental factors and maternal-infant
is difficult to distinguish from normal neonatal physiology after interactions could have been confounders, because AB+infants
birth, and laboratory tests cannot always reliably detect or rule were admitted to neonatal wards, whereas AB- infants stayed
out EOS.62 Because the consequences of delaying treatment are with their mothers on the maternity ward and were discharged
significant, on average 82 newborns without EOS are treated for earlier. Last, we did not have sufficient data on perinatal AB
each case.15 63–65 In our study population, only two of the 36 AB7 exposure and were therefore unable to correct for it, although
infants had a positive blood culture. The others were also treated it is questionable to what extent this confounder is important to
for 5–7 days because of elevated inflammatory markers or clin- take into account.75 76 Additionally, the study was not primarily
ical symptoms. Uzan-Yulzari et al showed that the association designed (and thus underpowered) to conclude on AB types.
between neonatal AB exposure and growth was independent of Nevertheless, our results suggest that ABAMX induced less pertur-
the neonatal infection state.6 This suggests that the differences bations as it did not result in any differences from AB- (online
in microbiota development after AB treatment in our study are supplemental figure 3B). The addition of the β-lactamase inhib-
more likely the result of the AB treatment duration itself than itor clavulanic acid (ABAMC) was associated with higher levels of
caused by a possible EOS. Our new findings emphasise the need bifidobacteria compared with other AB types, which supports
for microbiota restoration to minimise aberrant immune devel- an earlier finding in a single subject.20 Dedicated studies are,
opment. Suggested strategies include prebiotics, probiotics and however, needed to further elucidate the optimal regime with
synbiotics66 but also faecal transfers, which partially restored the the least microbial perturbations.
microbiota of mice exposed to AB for 7 days.12 In conclusion, AB exposure in the first week of life in term
In vaginally delivered infants, the AB effect was most born infants disturbed the microbiota up to 1 year, with more
F608 Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;107:F603–F610. doi:10.1136/archdischild-2021-322861
Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/archdischild-2021-322861 on 9 May 2022. Downloaded from http://fn.bmj.com/ on October 28, 2022 at World Health Organisation
Original research
revised the manuscript. JK is guarantor, supervised the research activity and 11 Kamphorst K, Van Daele E, Vlieger AM, et al. Early life antibiotics and childhood
revised the manuscript. gastrointestinal disorders: a systematic review. BMJ Paediatr Open 2021;5:e001028.
12 Niu X, Daniel S, Kumar D, et al. Transient neonatal antibiotic exposure increases
Funding This work was supported by JPI HDHL in conjunction with ZonMW and
susceptibility to late-onset sepsis driven by microbiota-dependent suppression of type
Danone Nutricia Research and grant number IM2015.
3 innate lymphoid cells. Sci Rep 2020;10:12974.
Competing interests This work and the PhD research by EVD was financed by a 13 Versporten A, Sharland M, Bielicki J, et al. The antibiotic resistance and prescribing
EU Joint Programming Initiative namely A Healthy Diet for a Healthy Life (JPIHDHL, in European children project: a neonatal and pediatric antimicrobial web-
http:// www.healthydietforhealthylife.eu/) inconjunction with ZonMW and Danone based point prevalence survey in 73 hospitals worldwide. Pediatr Infect Dis J
Nutricia Research. GH is a full full-time employee of Chr Hansen A/S since January 2013;32:e242–53.
2020. CB received a grant financed by JPI HDHL inconjunction with ZonMW and 14 Mukhopadhyay S, Eichenwald EC, Puopolo KM. Neonatal early-onset sepsis
Danone Nutricia Research. RMvE was an employee at Danone Nutricia Research till evaluations among well-appearing infants: projected impact of changes in CDC GBS
2020. JK is a full-time employee of Danone Nutricia. guidelines. J Perinatol 2013;33:198–205.
Patient consent for publication Consent obtained from parent(s)/guardian(s) 15 Fjalstad JW, Stensvold HJ, Bergseng H, et al. Early-Onset sepsis and antibiotic
exposure in term infants. Pediatr Infect Dis J 2016;35:1–6.
Ethics approval The INCA study was approved by the ethical board of the St. 16 Fjalstad JW, Esaiassen E, Juvet LK, et al. Antibiotic therapy in neonates and impact
Antonius Hospital in Nieuwegein, the Netherlands (registered as NCT02536560), on gut microbiota and antibiotic resistance development: a systematic review. J
and informed consent was obtained from both parents. Participants gave informed Antimicrob Chemother 2018;73:569–80.
consent to participate in the study before taking part. 17 Rooney AM, Timberlake K, Brown KA, et al. Each additional day of antibiotics is
Provenance and peer review Not commissioned; externally peer reviewed. associated with lower gut anaerobes in neonatal intensive care unit patients. Clin
Infect Dis 2020;70:2553–60.
Data availability statement Data are available on reasonable request. The data
18 Bennet R, Eriksson M, Nord CE. The fecal microflora of 1-3-month-old infants during
are available on request with the corresponding author.
treatment with eight oral antibiotics. Infection 2002;30:158-60.
Supplemental material This content has been supplied by the author(s). It 19 Parm U, Metsvaht T, Sepp E, et al. Impact of empiric antibiotic regimen on bowel
has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have colonization in neonates with suspected early onset sepsis. Eur J Clin Microbiol Infect
been peer-reviewed. Any opinions or recommendations discussed are solely those Dis 2010;29:807–16.
of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and 20 Korpela K, Salonen A, Saxen H. Antibiotics in early life associate with specific gut
responsibility arising from any reliance placed on the content. Where the content microbiota signatures in a prospective longitudinal infant cohort. Pediatr Res
includes any translated material, BMJ does not warrant the accuracy and reliability 2020:1–6.
of the translations (including but not limited to local regulations, clinical guidelines, 21 van den Anker J, Allegaert K. Rational use of antibiotics in neonates: still in search of
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;107:F603–F610. doi:10.1136/archdischild-2021-322861 F609
Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/archdischild-2021-322861 on 9 May 2022. Downloaded from http://fn.bmj.com/ on October 28, 2022 at World Health Organisation
Original research
43 Tanaka S, Kobayashi T, Songjinda P, et al. Influence of antibiotic exposure in the early 61 Benitz WE, Achten NB. Finding a role for the neonatal early-onset sepsis risk
postnatal period on the development of intestinal microbiota. FEMS Immunol Med calculator. EClinicalMedicine 2020;19:100255.
Microbiol 2009;56:80–7. 62 Sharma D, Farahbakhsh N, Shastri S. Biomarkers for diagnosis of neonatal sepsis:
44 Navarro-Tapia E, Sebastiani G, Sailer S, et al. Probiotic supplementation during a literature review;31:1646–1659, 2017. Available: https://doi.org/101080/
the perinatal and infant period: effects on gut dysbiosis and disease. Nutrients 1476705820171322060
2020;12:1–42. 63 Kerste M, Corver J, Sonnevelt MC. Application of sepsis calculator in newborns
45 Zwittink RD, Renes IB, van Lingen RA, et al. Association between duration of with suspected infection. 29:3860–3865, 2016. Available: https://doi.org/103109/
intravenous antibiotic administration and early-life microbiota development in late- 1476705820161149563
preterm infants. Eur J Clin Microbiol Infect Dis 2018;37:475–83. 64 van Herk W, Stocker M, van Rossum AMC. Recognising early onset neonatal sepsis: an
46 Lim HJ, Shin HS. Antimicrobial and Immunomodulatory Effects of Bifidobacterium essential step in appropriate antimicrobial use. J Infect 2016;72 Suppl:S77–82.
Strains: A Review. J Microbiol Biotechnol 2020;30:1793–800. 65 Goel N, Shrestha S, Smith R, et al. Screening for early onset neonatal sepsis: NICE
47 Croxen MA, Finlay BB. Molecular mechanisms of Escherichia coli pathogenicity. Nat guidance-based practice versus projected application of the Kaiser Permanente
Rev Microbiol 2010;8:26–38. sepsis risk calculator in the UK population. Arch Dis Child Fetal Neonatal Ed
48 Vatanen T, Kostic AD, d’Hennezel E, et al. Variation in microbiome LPS immunogenicity 2020;105:118–22.
contributes to autoimmunity in humans. Cell 2016;165:842–53. 66 Sohn K, Underwood MA. Prenatal and postnatal administration of prebiotics and
49 Zeevenhooven J, Browne PD, L’Hoir MP, L’Hoir MP, et al. Infant colic: mechanisms and probiotics. Semin Fetal Neonatal Med 2017;22:284–9.
management. Nat Rev Gastroenterol Hepatol 2018;15:479–96. 67 Bäckhed F, Roswall J, Peng Y, et al. Dynamics and stabilization of the human gut
50 Zimmermann P, Messina N, Mohn WW, et al. Association between the intestinal microbiome during the first year of life. Cell Host Microbe 2015;17:690–703.
microbiota and allergic sensitization, eczema, and asthma: A systematic review. J 68 Shao Y, Forster SC, Tsaliki E, et al. Stunted microbiota and opportunistic pathogen
Allergy Clin Immunol 2019;143:467–85. colonization in caesarean-section birth. Nature 2019;574:117–21.
51 de Weerth C, Fuentes S, Puylaert P, et al. Intestinal microbiota of infants with colic: 69 Martin R, Makino H, Cetinyurek Yavuz A, et al. Early-Life events, including mode
development and specific signatures. Pediatrics 2013;131:e550–8. of delivery and type of feeding, siblings and gender, shape the developing gut
52 Pittayanon R, Lau JT, Yuan Y, et al. Gut microbiota in patients with irritable bowel microbiota. PLoS One 2016;11:e0158498.
syndrome-a systematic review. Gastroenterology 2019;157:97–108. 70 Le Doare K, Holder B, Bassett A, et al. Mother’s milk: a purposeful contribution to the
53 Ramasethu J, Kawakita T. Antibiotic stewardship in perinatal and neonatal care. Semin development of the infant microbiota and immunity. Front Immunol 2018;9:361.
Fetal Neonatal Med 2017;22:278-283. 71 Rintala A, Pietilä S, Munukka E, et al. Gut microbiota analysis results are highly
54 Araujo da Silva AR, Albernaz de Almeida Dias DC, Marques AF, et al. Role of dependent on the 16S rRNA gene target region, whereas the impact of DNA
antimicrobial stewardship programmes in children: a systematic review. J Hosp Infect extraction is minor. J Biomol Tech 2017;28:19.
2018;99:117–23. 72 Chen Z, Hui PC, Hui M, et al. Impact of preservation method and 16S rRNA
55 Ibrahim NA, Makmor Bakry M, Ishak S, et al. A review of antibiotic used in suspected hypervariable region on gut microbiota profiling. mSystems 2019;4. doi:10.1128/
F610 Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;107:F603–F610. doi:10.1136/archdischild-2021-322861
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) Arch Dis Child Fetal Neonatal Ed
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;0:1–8. doi: 10.1136/archdischild-2021-322861
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) Arch Dis Child Fetal Neonatal Ed
Supplementary Figure 1
Supplementary Figure 1. Alpha diversity in age categories from birth to 2.5 years of age. Alpha diversity at different age categories
of the antibiotic (AB+) exposed and non-AB infants (AB-) did not differ at any of the time points, using one-way Analysis of
Covariance (ANCOVA) corrected for additional AB exposure between one and six months of age, ASV: Amplicon sequence variants.
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;0:1–8. doi: 10.1136/archdischild-2021-322861
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) Arch Dis Child Fetal Neonatal Ed
Supplementary Figure 2
Supplementary Figure 2. Unweighted UniFrac (UU) -based Principal Response Curves (PRC) complementary to the PRC in Figure
3. (a) The infants who did not receive antibiotics during the first week of life (AB-), were compared as a baseline to the antibiotic
exposed infants (AB+). Bacterial genera shown are the main drivers of the differences between AB+ and AB-: taxa on the same
side of baseline as the curve are linked to an increased relative abundance at that time point, opposite sides indicate a decrease
(b) AB- was also compared as a baseline with the different antibiotic durations of 2 to 3 (AB2) or 7 days (AB7) .
Significance was tested at the different time points using an ANOVA like permutation test (* = p-value < 0.05 compared to baseline
AB-). Covariates that were controlled for included additional AB exposure between the age of one and six months. AB-: infants
who did not receive AB during their first week of life, AB+: infants who received AB during their first week of life also indicated
within grey shading, AB2: AB exposure for 2 to 3 days in the first week of life, AB7: AB exposure for 7 days in the first week of life,
ASV: Amplicon sequence variants, UU: unweighted UniFrac.
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;0:1–8. doi: 10.1136/archdischild-2021-322861
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) Arch Dis Child Fetal Neonatal Ed
Supplementary Figure 3
Supplementary Figure 3. Weighted (WU) (a) and Unweighted UniFrac (UU) (b) -based Principal Response Curves (PRC) comparing
the impact of different antibiotic types administered during the first week of life. The infants who did not receive antibiotics during
the first week of life (AB-), were compared as a baseline to the different types of antibiotics ABPEN, ABAMX and ABAMC.
Significance was tested at the different time points using an ANOVA like permutation test (* = p-value < 0.05 compared to baseline
AB-). AB-: infants who did not receive AB during their first week of life, ABPEN: antibiotic exposure in first week of life with
gentamicin and penicillin, ABAMX: gentamicin and amoxicillin, ABAMC: gentamicin with amoxicillin and clavulanic acid. ASV:
Amplicon sequence variants, UU: unweighted UniFrac, WU: weighted UniFrac
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;0:1–8. doi: 10.1136/archdischild-2021-322861
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) Arch Dis Child Fetal Neonatal Ed
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;0:1–8. doi: 10.1136/archdischild-2021-322861
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance
Supplemental material placed on this supplemental material which has been supplied by the author(s) Arch Dis Child Fetal Neonatal Ed
1. Hoftiezer, L. et al. From population reference to national standard: new and improved
birthweight charts. Am. J. Obstet. Gynecol. 220, 383.e1-383.e17 (2019).
2. Milani, C. et al. Assessing the Fecal Microbiota: An Optimized Ion Torrent 16S rRNA Gene-
Based Analysis Protocol. PLoS One 8, e68739 (2013).
Van Daele E, et al. Arch Dis Child Fetal Neonatal Ed 2022;0:1–8. doi: 10.1136/archdischild-2021-322861